********************************************************************************************************************************
*** REPLICATION MATERIALS FOR ARTICLE, "DO MINIMUM CHARITY CARE POLICY REQUIREMENTS INCREASE NONPROFIT HOSPITAL PERFORMANCE? ***
********************************************************************************************************************************

** Read data
use "C:\Users\great\Atlas of Collab Dropbox\NARA YOON\04. JOURNAL SUBMISSION\2020\MCCP R&R\JPART\03. FINAL SUBMISSION FOR PUBLICATION\Data for replication purpose", clear



****************************************************************************************************************************
** Table 1. Mean Hospital and County Characteristics, Hospitals Continuously Operating 2009-2015, Characteristics in 2009 **
****************************************************************************************************************************

** Hospital and patient characteristics
tabstat real_rev_total n_all CONbed if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==1, format(%9.0f)
tabstat Pct_Charity Pct_rev_Charity if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==1, s(mean sd) format(%9.2f) 

tabstat real_rev_total n_all CONbed if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==1, by(type_new) format(%9.0f)
tabstat Pct_Charity Pct_rev_Charity if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==1, s(mean sd) by(type_new) format(%9.2f)

tabstat real_rev_total n_all CONbed if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==0, format(%9.0f)
tabstat Pct_Charity Pct_rev_Charity if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==0, s(mean sd) format(%9.2f) 


** Number of hospitals
unique encode_id if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==1  
unique encode_id if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==1 & NP==1
unique encode_id if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==1 & FP==1
unique encode_id if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==1 & GOV==1
unique encode_id if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==0


** County characteristics
tabstat ACS_POP real_ACS_CAPITA_INCOME if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==1, format(%9.0f)
tabstat ACS_R_FEMALE ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_WHITE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_HISPANIC ACS_R_BORN_OUTSIDEUS ACS_R_ENGLISH_LESS ACS_R_EDU_LESSHIGH ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L1 ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==1, format(%9.2f)

tabstat ACS_POP real_ACS_CAPITA_INCOME if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==1, by(type_new) format(%9.0f)
tabstat ACS_R_FEMALE ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_WHITE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_HISPANIC ACS_R_BORN_OUTSIDEUS ACS_R_ENGLISH_LESS ACS_R_EDU_LESSHIGH ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L1 ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==1, by(type_new) format(%9.2f)

tabstat ACS_POP real_ACS_CAPITA_INCOME if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==0, format(%9.0f)
tabstat ACS_R_FEMALE ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_WHITE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_HISPANIC ACS_R_BORN_OUTSIDEUS ACS_R_ENGLISH_LESS ACS_R_EDU_LESSHIGH ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L1 ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==0, format(%9.2f)


** Number of counties
unique cty2 if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==1 
unique cty2 if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==1 & NP==1
unique cty2 if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==1 & FP==1
unique cty2 if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==1 & GOV==1
unique cty2 if continuous_operate==1 & analytic_sample==1 & year==2009 & auditgen==0



**************************************************************************************************************************
** Table 2. Regression Results, Impact of Minimum Charity Care Requirements, Difference-in-Differences Model, 2009-2015 **
**************************************************************************************************************************

** DV1: Percentage of patients receiving charity care 
* 1) year fe (YES), county control (NO), hospital control(NO), county fe (NO), hospital fe (NO)
reg Pct_Charity NPPost NP FP yr10 yr11 yr12 yr13 yr14 yr15 if analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* 2) year fe (YES), county control (YES), hospital control(NO), county fe (NO), hospital fe (NO)
reg Pct_Charity NPPost NP FP yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65  ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS if analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* 3) year fe (YES), county control (YES), hospital control(YES), county fe (NO), hospital fe (NO)
reg Pct_Charity NPPost NP FP yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65  ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS real_l_rev_total CONbed if analytic_sample==1 & auditgen==1 & continuous_operate==1 & continuous_operate==1, cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* 4) year fe (YES), county control (YES), hospital control(YES), county fe (YES), hospital fe (NO)
areg Pct_Charity NPPost NP FP yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65  ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS real_l_rev_total CONbed if analytic_sample==1 & auditgen==1 & continuous_operate==1, absorb (cty2) cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* 5) year fe (YES), county control (YES), hospital control(YES), county fe (NO), hospital fe (YES)
areg Pct_Charity NPPost NP FP yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65  ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS real_l_rev_total if analytic_sample==1 & auditgen==1 & continuous_operate==1, absorb (encode_id) cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1



** DV2: Charity care spending relative to client revenues earned 
* 1) year fe (YES), county control (NO), hospital control(NO), county fe (NO), hospital fe (NO)
reg Pct_rev_Charity NPPost NP FP yr10 yr11 yr12 yr13 yr14 yr15 if analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* 2) year fe (YES), county control (YES), hospital control(NO), county fe (NO), hospital fe (NO)
reg Pct_rev_Charity NPPost NP FP yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65  ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS if analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* 3) year fe (YES), county control (YES), hospital control(YES), county fe (NO), hospital fe (NO)
reg Pct_rev_Charity NPPost NP FP yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65  ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS real_l_rev_total CONbed if analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* 4) year fe (YES), county control (YES), hospital control(YES), county fe (YES), hospital fe (NO)
areg Pct_rev_Charity NPPost NP FP yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65  ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS real_l_rev_total CONbed if analytic_sample==1 & auditgen==1 & continuous_operate==1, absorb (cty2) cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* 5) year fe (YES), county control (YES), hospital control(YES), county fe (NO), hospital fe (YES)
areg Pct_rev_Charity NPPost NP FP yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65  ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS real_l_rev_total if analytic_sample==1 & auditgen==1 & continuous_operate==1, absorb (encode_id) cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1



***************************************************************************
** Table 3. Heterogeneity of Impacts by Baseline Charity Care Provisions **
***************************************************************************

** DV1: Percentage of patients receiving charity care
* 1) year fe (YES), county control (NO), hospital control(NO), county fe (NO), hospital fe (NO)
reg Pct_Charity  PFirst_Pct_rev_Charity_terciles PSecond_Pct_rev_Charity_terciles PThird_Pct_rev_Charity_terciles First_Pct_rev_Charity_terciles Second_Pct_rev_Charity_terciles Third_Pct_rev_Charity_terciles FP09 yr10 yr11 yr12 yr13 yr14 yr15 if analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust

test PFirst_Pct_rev_Charity_terciles = PSecond_Pct_rev_Charity_terciles
test PFirst_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles
test PSecond_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles

unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* 2) year fe (YES), county control (YES), hospital control(NO), county fe (NO), hospital fe (NO)
reg Pct_Charity PFirst_Pct_rev_Charity_terciles PSecond_Pct_rev_Charity_terciles PThird_Pct_rev_Charity_terciles First_Pct_rev_Charity_terciles Second_Pct_rev_Charity_terciles Third_Pct_rev_Charity_terciles FP09 yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS if analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust

test PFirst_Pct_rev_Charity_terciles = PSecond_Pct_rev_Charity_terciles
test PFirst_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles
test PSecond_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles

unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* 3) year fe (YES), county control (YES), hospital control(YES), county fe (NO), hospital fe (NO)
reg Pct_Charity PFirst_Pct_rev_Charity_terciles PSecond_Pct_rev_Charity_terciles PThird_Pct_rev_Charity_terciles First_Pct_rev_Charity_terciles Second_Pct_rev_Charity_terciles Third_Pct_rev_Charity_terciles FP09 yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS real_l_rev_total CONbed if analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust

test PFirst_Pct_rev_Charity_terciles = PSecond_Pct_rev_Charity_terciles
test PFirst_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles
test PSecond_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles

unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* 4) year fe (YES), county control (YES), hospital control(YES), county fe (YES), hospital fe (NO)
areg Pct_Charity PFirst_Pct_rev_Charity_terciles PSecond_Pct_rev_Charity_terciles PThird_Pct_rev_Charity_terciles First_Pct_rev_Charity_terciles Second_Pct_rev_Charity_terciles Third_Pct_rev_Charity_terciles FP09 yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS real_l_rev_total CONbed if analytic_sample==1 & auditgen==1 & continuous_operate==1, absorb (cty2) cluster (encode_id) robust

test PFirst_Pct_rev_Charity_terciles = PSecond_Pct_rev_Charity_terciles
test PFirst_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles
test PSecond_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles

unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* 5) year fe (YES), county control (YES), hospital control(YES), county fe (NO), hospital fe (YES)
areg Pct_Charity PFirst_Pct_rev_Charity_terciles PSecond_Pct_rev_Charity_terciles PThird_Pct_rev_Charity_terciles yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS real_l_rev_total if analytic_sample==1 & auditgen==1 & continuous_operate==1, absorb (encode_id) cluster (encode_id) robust

test PFirst_Pct_rev_Charity_terciles = PSecond_Pct_rev_Charity_terciles
test PFirst_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles
test PSecond_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles

unique cty2 if e(sample)==1
unique encode_id if e(sample)==1



** DV 2: Charity care spending relative to client revenues earned
* 1) year fe (YES), county control (NO), hospital control(NO), county fe (NO), hospital fe (NO)
reg Pct_rev_Charity PFirst_Pct_rev_Charity_terciles PSecond_Pct_rev_Charity_terciles PThird_Pct_rev_Charity_terciles First_Pct_rev_Charity_terciles Second_Pct_rev_Charity_terciles Third_Pct_rev_Charity_terciles FP09 yr10 yr11 yr12 yr13 yr14 yr15 if analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust

test PFirst_Pct_rev_Charity_terciles = PSecond_Pct_rev_Charity_terciles
test PFirst_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles
test PSecond_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles

unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* 2) year fe (YES), county control (YES), hospital control(NO), county fe (NO), hospital fe (NO)
reg Pct_rev_Charity PFirst_Pct_rev_Charity_terciles PSecond_Pct_rev_Charity_terciles PThird_Pct_rev_Charity_terciles First_Pct_rev_Charity_terciles Second_Pct_rev_Charity_terciles Third_Pct_rev_Charity_terciles FP09 yr10 yr11 yr12 yr13 yr14 yr15  ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS if analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust

test PFirst_Pct_rev_Charity_terciles = PSecond_Pct_rev_Charity_terciles
test PFirst_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles
test PSecond_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles

unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* 3) year fe (YES), county control (YES), hospital control(YES), county fe (NO), hospital fe (NO)
reg Pct_rev_Charity PFirst_Pct_rev_Charity_terciles PSecond_Pct_rev_Charity_terciles PThird_Pct_rev_Charity_terciles First_Pct_rev_Charity_terciles Second_Pct_rev_Charity_terciles Third_Pct_rev_Charity_terciles FP09 yr10 yr11 yr12 yr13 yr14 yr15  ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS real_l_rev_total CONbed if analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust

test PFirst_Pct_rev_Charity_terciles = PSecond_Pct_rev_Charity_terciles
test PFirst_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles
test PSecond_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles

unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* 4) year fe (YES), county control (YES), hospital control(YES), county fe (YES), hospital fe (NO)
areg Pct_rev_Charity PFirst_Pct_rev_Charity_terciles PSecond_Pct_rev_Charity_terciles PThird_Pct_rev_Charity_terciles First_Pct_rev_Charity_terciles Second_Pct_rev_Charity_terciles Third_Pct_rev_Charity_terciles FP09 yr10 yr11 yr12 yr13 yr14 yr15  ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS real_l_rev_total CONbed if analytic_sample==1 & auditgen==1 & continuous_operate==1, absorb (cty2) cluster (encode_id) robust

test PFirst_Pct_rev_Charity_terciles = PSecond_Pct_rev_Charity_terciles
test PFirst_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles
test PSecond_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles

unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* 5) year fe (YES), county control (YES), hospital control(YES), county fe (NO), hospital fe (YES)
areg Pct_rev_Charity PFirst_Pct_rev_Charity_terciles PSecond_Pct_rev_Charity_terciles PThird_Pct_rev_Charity_terciles yr10 yr11 yr12 yr13 yr14 yr15  ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS real_l_rev_total if analytic_sample==1 & auditgen==1 & continuous_operate==1, absorb (encode_id) cluster (encode_id) robust

test PFirst_Pct_rev_Charity_terciles = PSecond_Pct_rev_Charity_terciles
test PFirst_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles
test PSecond_Pct_rev_Charity_terciles = PThird_Pct_rev_Charity_terciles

unique cty2 if e(sample)==1
unique encode_id if e(sample)==1





*************************************************************************************
** Appendix 1. Share of Hospital by Institutional Form, Analytic Sample, 2009-2015 **
*************************************************************************************
sort hospitalid year
tab type if continuous_operate==1 & analytic_sample==1 & auditgen==1 & year==2009
tab type if continuous_operate==1 & analytic_sample==1 & auditgen==1 & year==2010
tab type if continuous_operate==1 & analytic_sample==1 & auditgen==1 & year==2011
tab type if continuous_operate==1 & analytic_sample==1 & auditgen==1 & year==2012
tab type if continuous_operate==1 & analytic_sample==1 & auditgen==1 & year==2013
tab type if continuous_operate==1 & analytic_sample==1 & auditgen==1 & year==2014
tab type if continuous_operate==1 & analytic_sample==1 & auditgen==1 & year==2015



*******************************************************************************************************************************
** Appendix 2. Regression Results, Relationship Between Institutional Form and Charity Care Provision, Pre-Policy, 2009-2011 **
*******************************************************************************************************************************

** DV1: Percentage of patients receiving charity care
* column 1) year fe (YES), county fe(NO), county control (NO), hospital control(NO)
reg Pct_Charity NP FP yr10 yr11 if year <2012 & analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust
test NP=FP
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1

* column 2) year fe (YES), county control (YES), hospital control(NO), county fe(NO)
reg Pct_Charity NP FP yr10 yr11 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS if year <2012 & analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust
test NP=FP
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1

* column 3) year fe (YES), county control (YES), hospital control(YES), county fe(NO)
reg Pct_Charity NP FP yr10 yr11 real_l_rev_total CONbed ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65  ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS if year <2012 & analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust
test NP=FP
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1

* column 4) year fe (YES), county control (YES), hospital control(YES), county fe(YES)
areg Pct_Charity NP FP yr10 yr11 real_l_rev_total CONbed ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65  ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS if year <2012 & analytic_sample==1 & auditgen==1 & continuous_operate==1, absorb (cty2) cluster (encode_id) robust
test NP=FP
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1



** DV 2: Charity care spending relative to client revenues earned
* column 5) year fe (YES), county control (NO), hospital control(NO), county fe(NO)
reg Pct_rev_Charity NP FP yr10 yr11 if year <2012 & analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust
test NP=FP
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1

* column 6) year fe (YES), county control (YES), hospital control(NO), county fe(NO)
reg Pct_rev_Charity NP FP yr10 yr11 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS if year <2012 & analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust
test NP=FP
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1

* column 7) year fe (YES), county control (YES), hospital control(YES), county fe(NO)
reg Pct_rev_Charity NP FP yr10 yr11 real_l_rev_total CONbed ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65  ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS if year <2012 & analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust
test NP=FP
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* column 8) year fe (YES), county control (YES), hospital control(YES), county fe(YES)
areg Pct_rev_Charity NP FP yr10 yr11 real_l_rev_total CONbed ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65  ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS if year <2012 & analytic_sample==1 & auditgen==1 & continuous_operate==1, absorb (cty2) cluster (encode_id) robust
test NP=FP
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1




***********************************************************************************************************************************
** Appendix 3. Test of Parallel Trend Assumption (Pre-Policy, 2009-2011) and Placebo Test (Impact Year Before Policy), 2009-2015 **
***********************************************************************************************************************************

** DV1: Percentage of patients receiving charity care

* column 1) year fe (YES), county control (YES), hospital control(YES), county fe (NO)
reg Pct_Charity NPPretrend NP FP yr10 yr11 real_l_rev_total CONbed ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65  ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS if year <2012 & analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* column 2) year fe (YES), county control (YES), hospital control(YES), county fe (YES) -- MAY NEED TO DROP COUNTY FE DUE TO SMALL VARIATIONS! MISSING F-STATS
areg Pct_Charity NPPretrend NP FP yr10 yr11 real_l_rev_total CONbed ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65  ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS if year <2012 & analytic_sample==1 & auditgen==1 & continuous_operate==1, absorb (cty2) cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* column 3) year fe (YES), county control (YES), hospital control(YES), county fe (NO), hospital fe (NO)
reg Pct_Charity NPfakePost NPPost NP FP yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC   ACS_R_ENGLISH_LESS  real_l_rev_total CONbed if analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* column 4) year fe (YES), county control (YES), hospital control(YES), county fe (YES), hospital fe (NO)
areg Pct_Charity NPfakePost NPPost NP FP yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC   ACS_R_ENGLISH_LESS real_l_rev_total CONbed if analytic_sample==1 & auditgen==1 & continuous_operate==1, absorb (cty2) cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* column 5) year fe (YES), county control (YES), hospital control(YES), county fe (NO), hospital fe (YES)
areg Pct_Charity NPfakePost NPPost NP FP yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC  ACS_R_ENGLISH_LESS  real_l_rev_total if analytic_sample==1 & auditgen==1 & continuous_operate==1, absorb (encode_id) cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1




** DV 2: Charity care spending relative to client revenues earned 

* column 6) year fe (YES), county control (YES), hospital control(YES), county fe (NO)
reg Pct_rev_Charity NPPretrend NP FP yr10 yr11 real_l_rev_total CONbed ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65  ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS if year <2012 & analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1

* column 7) year fe (YES), county control (YES), hospital control(YES), county fe (YES)-- MAY NEED TO DROP COUNTY FE DUE TO SMALL VARIATIONS! MISSING F-STATS
areg Pct_rev_Charity NPPretrend NP FP yr10 yr11 real_l_rev_total CONbed ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65  ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS if year <2012 & analytic_sample==1 & auditgen==1 & continuous_operate==1, absorb (cty2) cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* column 8) year fe (YES), county control (YES), hospital control(YES), county fe (NO), hospital fe (NO)
reg Pct_rev_Charity NPfakePost NPPost NP FP yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS  real_l_rev_total CONbed if analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* column 9) year fe (YES), county control (YES), hospital control(YES), county fe (YES), hospital fe (NO)
areg Pct_rev_Charity NPfakePost NPPost NP FP yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS  real_l_rev_total CONbed if analytic_sample==1 & auditgen==1 & continuous_operate==1, absorb (cty2) cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


* column 10) year fe (YES), county control (YES), hospital control(YES), county fe (NO), hospital fe (YES)
areg Pct_rev_Charity NPfakePost NPPost NP FP yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS real_l_rev_total if analytic_sample==1 & auditgen==1 & continuous_operate==1, absorb (encode_id) cluster (encode_id) robust
unique cty2 if e(sample)==1
unique encode_id if e(sample)==1


 
 
***************************************************
** Appendix 4. Panel A-B. Event Study, 2009-2015 **
***************************************************

** Create baseline group 
gen yr11_NP_Zero = 0

** DV1: Percentage of patients receiving charity care
* year fe (YES), county control (YES), hospital control(YES), county fe (NO), hospital fe (NO)
eststo Z_Mathall_Pct_Charity: reg Pct_Charity yr09_NP yr10_NP yr11_NP_Zero yr12_NP yr13_NP yr14_NP yr15_NP NP FP yr09 yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS real_l_rev_total CONbed if analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust

coefplot (Z_Mathall_Pct_Charity), drop (_cons NP FP yr09 yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS real_l_rev_total CONbed) groups (yr09_NP yr10_NP yr11_NP_Zero yr12_NP yr13_NP yr14_NP yr15_NP = "Year")  coeflabels( yr09_NP= "2009" yr10_NP= "2010" yr11_NP_Zero="2011" yr12_NP= "2012" yr13_NP = "2013" yr14_NP = "2014" yr15_NP = "2015", wrap(4)) ylabel(-6(2)2) yscale(range(-6 2)) yline(0) vertical ytitle ("Percentage of Patients Receiving Charity Care", size(small)) omitted



** DV 2: Charity care spending relative to client revenues earned
* year fe (YES), county control (YES), hospital control(YES), county fe (NO), hospital fe (NO)
eststo Z_Mathall_Pct_rev_Charity: reg Pct_rev_Charity yr09_NP yr10_NP yr11_NP_Zero yr12_NP yr13_NP yr14_NP yr15_NP NP FP yr09 yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS real_l_rev_total CONbed if analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust

coefplot (Z_Mathall_Pct_rev_Charity), drop (_cons NP FP yr09 yr10 yr11 yr12 yr13 yr14 yr15 ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS real_l_rev_total CONbed) groups (yr09_NP yr10_NP yr11_NP_Zero yr12_NP yr13_NP yr14_NP yr15_NP = "Year")  coeflabels( yr09_NP= "2009" yr10_NP= "2010" yr11_NP_Zero="2011" yr12_NP= "2012" yr13_NP = "2013" yr14_NP = "2014" yr15_NP = "2015", wrap(4)) ylabel(-6(2)2) yscale(range(-6 2)) yline(0) vertical ytitle ("Charity Care Spending Relative to Client Revenues Earned", size(small)) omitted



***************************************************************************************************************
** Appendix 5. Panel A-B. Event Study Results, Heterogeneity of Impacts by Baseline Charity Care, 2009-2015  **
***************************************************************************************************************

** Create baseline group 
gen yr11_NP09_mid_rev_Zero=0 

** DV1: Percentage of patients receiving charity care
* year fe (YES), county control (YES), hospital control(YES), county fe (NO), hospital fe (NO)
reg Pct_Charity yr09_NP09_top_rev yr10_NP09_top_rev yr11_NP09_top_rev yr12_NP09_top_rev yr13_NP09_top_rev yr14_NP09_top_rev yr15_NP09_top_rev ///
yr09_NP09_mid_rev yr10_NP09_mid_rev yr11_NP09_mid_rev_Zero yr12_NP09_mid_rev yr13_NP09_mid_rev yr14_NP09_mid_rev yr15_NP09_mid_rev ///
yr09_NP09_bot_rev yr10_NP09_bot_rev yr11_NP09_bot_rev yr12_NP09_bot_rev yr13_NP09_bot_rev yr14_NP09_bot_rev yr15_NP09_bot_rev ///
FP09 NP09 yr09 yr10 yr12 yr13 yr14 yr15 ///
ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER    ///
ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT ///
real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS ///
real_l_rev_total CONbed       ///
if analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust

gen coeff_top_rev=.
gen coeff_mid_rev=.
gen coeff_bot_rev=.

replace coeff_top_rev = _b[yr09_NP09_top_rev] if year==2009 & First_Pct_rev_Charity_terciles==1 
replace coeff_mid_rev = _b[yr09_NP09_mid_rev] if year==2009 & Second_Pct_rev_Charity_terciles==1 
replace coeff_bot_rev = _b[yr09_NP09_bot_rev] if year==2009 & Third_Pct_rev_Charity_terciles==1 

replace coeff_top_rev = _b[yr10_NP09_top_rev] if year==2010 & First_Pct_rev_Charity_terciles==1 
replace coeff_mid_rev = _b[yr10_NP09_mid_rev] if year==2010 & Second_Pct_rev_Charity_terciles==1 
replace coeff_bot_rev = _b[yr10_NP09_bot_rev] if year==2010 & Third_Pct_rev_Charity_terciles==1 

replace coeff_top_rev = _b[yr11_NP09_top_rev] if year==2011 & First_Pct_rev_Charity_terciles==1 
replace coeff_mid_rev = _b[yr11_NP09_mid_rev_Zero] if year==2011 & Second_Pct_rev_Charity_terciles==1 
replace coeff_bot_rev = _b[yr11_NP09_bot_rev] if year==2011 & Third_Pct_rev_Charity_terciles==1 

replace coeff_top_rev = _b[yr12_NP09_top_rev] if year==2012 & First_Pct_rev_Charity_terciles==1 
replace coeff_mid_rev = _b[yr12_NP09_mid_rev] if year==2012 & Second_Pct_rev_Charity_terciles==1 
replace coeff_bot_rev = _b[yr12_NP09_bot_rev] if year==2012 & Third_Pct_rev_Charity_terciles==1 

replace coeff_top_rev = _b[yr13_NP09_top_rev] if year==2013 & First_Pct_rev_Charity_terciles==1 
replace coeff_mid_rev = _b[yr13_NP09_mid_rev] if year==2013 & Second_Pct_rev_Charity_terciles==1 
replace coeff_bot_rev = _b[yr13_NP09_bot_rev] if year==2013 & Third_Pct_rev_Charity_terciles==1 

replace coeff_top_rev = _b[yr14_NP09_top_rev] if year==2014 & First_Pct_rev_Charity_terciles==1 
replace coeff_mid_rev = _b[yr14_NP09_mid_rev] if year==2014 & Second_Pct_rev_Charity_terciles==1 
replace coeff_bot_rev = _b[yr14_NP09_bot_rev] if year==2014 & Third_Pct_rev_Charity_terciles==1 

replace coeff_top_rev = _b[yr15_NP09_top_rev] if year==2015 & First_Pct_rev_Charity_terciles==1 
replace coeff_mid_rev = _b[yr15_NP09_mid_rev] if year==2015 & Second_Pct_rev_Charity_terciles==1 
replace coeff_bot_rev = _b[yr15_NP09_bot_rev] if year==2015 & Third_Pct_rev_Charity_terciles==1 

twoway (connected coeff_top_rev year, sort lpattern(solid) msymbol(S) xlabel(2009(1)2015) ylabel(-2.5(1)2.5) ytitle("Percentage of Patients Receiving Charity Care", size(small))) (connected coeff_mid_rev year, sort lpattern(longdash) msymbol(X) xlabel(2009(1)2015) ylabel(-2.5(1)2.5) yline(0) xline(2011.5, lcolor(black) lpattern(shortdash))) (connected coeff_bot_rev year, sort lpattern(shortdash_dot) msymbol(T) xlabel(2009(1)2015) ylabel(-2.5(1)2.5)) if year>=2009 & year<=2015, legend(label(1 "Top") label(2 "Middle") label(3 "Bottom"))


** DV 2: Charity care spending relative to client revenues earned 
* year fe (YES), county control (YES), hospital control(YES), county fe (NO), hospital fe (NO)
reg Pct_rev_Charity yr09_NP09_top_rev yr10_NP09_top_rev yr11_NP09_top_rev yr12_NP09_top_rev yr13_NP09_top_rev yr14_NP09_top_rev yr15_NP09_top_rev ///
yr09_NP09_mid_rev yr10_NP09_mid_rev yr11_NP09_mid_rev_Zero yr12_NP09_mid_rev yr13_NP09_mid_rev yr14_NP09_mid_rev yr15_NP09_mid_rev ///
yr09_NP09_bot_rev yr10_NP09_bot_rev yr11_NP09_bot_rev yr12_NP09_bot_rev yr13_NP09_bot_rev yr14_NP09_bot_rev yr15_NP09_bot_rev ///
FP09 NP09 yr09 yr10 yr12 yr13 yr14 yr15 ///
ACS_POP ACS_R_UNDER_18 ACS_R_OVER_65 ACS_R_FEMALE ACS_R_BLACK ACS_R_ALL_OTHER    ///
ACS_R_EDU_HIGH ACS_R_EDU_SOMECOLLEGE ACS_R_EDU_BACHELOR ACS_R_EDU_GRAD ACS_R_POVERTY_L2 ACS_R_POVERTY_L3 ACS_R_UNEMPLOYMENT ///
real_ACS_CAPITA_INCOME ACS_R_BORN_OUTSIDEUS ACS_R_HISPANIC ACS_R_ENGLISH_LESS ///
real_l_rev_total CONbed       ///
if analytic_sample==1 & auditgen==1 & continuous_operate==1, cluster (encode_id) robust


gen coeff_top_rev2=.
gen coeff_mid_rev2=.
gen coeff_bot_rev2=.

replace coeff_top_rev2 = _b[yr09_NP09_top_rev] if year==2009 & First_Pct_rev_Charity_terciles==1 
replace coeff_mid_rev2 = _b[yr09_NP09_mid_rev] if year==2009 & Second_Pct_rev_Charity_terciles==1 
replace coeff_bot_rev2 = _b[yr09_NP09_bot_rev] if year==2009 & Third_Pct_rev_Charity_terciles==1 

replace coeff_top_rev2 = _b[yr10_NP09_top_rev] if year==2010 & First_Pct_rev_Charity_terciles==1 
replace coeff_mid_rev2 = _b[yr10_NP09_mid_rev] if year==2010 & Second_Pct_rev_Charity_terciles==1 
replace coeff_bot_rev2 = _b[yr10_NP09_bot_rev] if year==2010 & Third_Pct_rev_Charity_terciles==1 

replace coeff_top_rev2 = _b[yr11_NP09_top_rev] if year==2011 & First_Pct_rev_Charity_terciles==1 
replace coeff_mid_rev2 = _b[yr11_NP09_mid_rev_Zero] if year==2011 & Second_Pct_rev_Charity_terciles==1 
replace coeff_bot_rev2 = _b[yr11_NP09_bot_rev] if year==2011 & Third_Pct_rev_Charity_terciles==1 

replace coeff_top_rev2 = _b[yr12_NP09_top_rev] if year==2012 & First_Pct_rev_Charity_terciles==1 
replace coeff_mid_rev2 = _b[yr12_NP09_mid_rev] if year==2012 & Second_Pct_rev_Charity_terciles==1 
replace coeff_bot_rev2 = _b[yr12_NP09_bot_rev] if year==2012 & Third_Pct_rev_Charity_terciles==1 

replace coeff_top_rev2 = _b[yr13_NP09_top_rev] if year==2013 & First_Pct_rev_Charity_terciles==1 
replace coeff_mid_rev2 = _b[yr13_NP09_mid_rev] if year==2013 & Second_Pct_rev_Charity_terciles==1 
replace coeff_bot_rev2 = _b[yr13_NP09_bot_rev] if year==2013 & Third_Pct_rev_Charity_terciles==1 

replace coeff_top_rev2 = _b[yr14_NP09_top_rev] if year==2014 & First_Pct_rev_Charity_terciles==1 
replace coeff_mid_rev2 = _b[yr14_NP09_mid_rev] if year==2014 & Second_Pct_rev_Charity_terciles==1 
replace coeff_bot_rev2 = _b[yr14_NP09_bot_rev] if year==2014 & Third_Pct_rev_Charity_terciles==1 

replace coeff_top_rev2 = _b[yr15_NP09_top_rev] if year==2015 & First_Pct_rev_Charity_terciles==1 
replace coeff_mid_rev2 = _b[yr15_NP09_mid_rev] if year==2015 & Second_Pct_rev_Charity_terciles==1 
replace coeff_bot_rev2 = _b[yr15_NP09_bot_rev] if year==2015 & Third_Pct_rev_Charity_terciles==1 

twoway (connected coeff_top_rev2 year, sort lpattern(solid) msymbol(S) xlabel(2009(1)2015) ylabel(-2.5(1)2.5) ytitle("Charity Care Spending Relative to Client Revenues Earned", size(small))) (connected coeff_mid_rev2 year, sort lpattern(longdash) msymbol(X) xlabel(2009(1)2015) ylabel(-2.5(1)2.5)  yline(0) xline(2011.5, lcolor(black) lpattern(shortdash))) (connected coeff_bot_rev2 year, sort lpattern(shortdash_dot) msymbol(T) xlabel(2009(1)2015) ylabel(-2.5(1)2.5))  if year>=2009 & year<=2015, legend(label(1 "Top") label(2 "Middle") label(3 "Bottom"))



